Re: [Tracker] Tracker only indexing odt Heading contents

[Karl Relton <karllinuxtest relton ntlworld com>]

On Fri, 2012-04-06 at 15:43 +0100, Karl Relton wrote:
> Hi all
>
> Trackers odt extraction doesn't seem to be the full text. From what I
> can see, it is only storing words in paragraphs using 'heading' styles
> (any of 'heading' of 'heading1', 'heading2' etc.). Text in other
> paragraphs (e.g. using 'Default' or 'Text body' is not being indexed.
>
> Is this intentional?

Looking at the code in tracker-extract-oasis.c, it seems unnecessarily
picky to me.

For the main text content (ODF elements text:p & text:h) it is looking
at whether styles are 'present', or are certain types. Why this effort?
Why not just accept all content in text:p or text:h elements, regardless
of styles etc.? That would seem to me to be more in line with the ODF
spec.


Also, as an aside, it seems pointless capturing stuff in elements
text:s, since this is by definition just white space.